PDA

View Full Version : 3D Printer questions



Evan
06-05-2013, 02:43 PM
I would like to add a 3D printer head and extruder etc to my milling machine as a removable accessory. I spent some time looking at what is available online in terms of plans and what I mostly find are various ways to minimize the amount of machining and the tools required to build something. That is not an issue for me as I have the capability to build virtually any complex design. What I would like is some clue to what would be the best design that will take best advantage of the CNC mill I already have. Once I have that running I then may build a stand alone machine if I find additive machining interesting enough using just plastics.

Anybody know what I should be looking for in order to build an add on additive machining head to my mill? I know almost nothing about the actual required parts and pieces in this field. All I want are the specs/plans/requirements to build what I need. I don't need actual plans if the necessary specs are good enough. However, there is no point in reinventing the wheel so I will happily go with somebody else's proven design.

small.planes
06-05-2013, 02:56 PM
Id look at the Bowden Cable remote motor type extruders.
You can mount the moter / drive unit mostly anywhere, and only have to move the hot end.

Which plastics are you looking at using?

Here:http://reprap.org/wiki/Category:Extruders is as good a place to start as any
Mostly the plans are available in a GIT repositry

Dave

Evan
06-05-2013, 03:42 PM
The main thing I am looking for is designs where there are not compromises made to accommodate lack of machining tools. I would rather make my own unless there is something that I cannot feasibly make as well as a "store bought" unit, such as a heating element etc.

As for plastics, most likely ABS unless there is a better but still reasonably priced alternative. DX sells ABS spools at around $45 for a kilo in various colours. I have no idea if that is a good price or not.

I am open to all ideas and advice. This is an area I have not yet explored (much).

aostling
06-05-2013, 04:00 PM
Because of its high resolution at (relatively) low cost, I've been waiting for the Form 1 to become widely available http://formlabs.com/products/our-printer. Instead of buying one initially I'd like to send a file to a service which could print a part for my inspection. I hope such a service will become available, but I have no knowledge of any.

Evan
06-05-2013, 04:29 PM
I wonder what would happen if you used a MIG welder to build up metal with tiny "spot welds", using CNC timing and position control?

I see that this was mentioned here before but with only one reply it didn't go far. http://bbs.homeshopmachinist.net/threads/55863-Tig-welder-3D-printer

I don't want to sidetrack my own thread though. If anybody has knowledge to share about plastic printing I am all "ears".

small.planes
06-05-2013, 04:41 PM
Bear in mind Ive personally only printed 5 parts, but I have been using these technologies for 4or so years.
That out of the way onto the other words.
ABS is prone to warping, and all the home printers use a heated printing surface to help, but it apparently doesnt always work.
If PLA will suit its much more forgiving to print with.

The extruders are all pretty basic tech. Use a stepper to push the filament a known distance (and therefore a known volume) through a heated tube with a 0.3mm hole in the end and move the nozzle to print the shape.
The devil is in the detail of course...
Keeping the transition between solid and melted as short as possible apparently assists in accurate placement and volume of plastic delivered.
Temperature control is also reasonably critical as far as I can see, overheating will ruin the plastic, underheating makes the push harder to extrude.
I dont think that all the extruder designs are compromised by lack of machining ability, they are more designed in a different paradigm

Dave

small.planes
06-05-2013, 04:44 PM
The problem with mig printing metal is the high contraction rate, causing warping. I did some work with some research people into this as an alternative to sls powder metalurgy. As far as I know it still doesnt work properly, even in a heated argon chamber to control cooling rates.

Dave

JohnAlex141r
06-05-2013, 04:55 PM
Evan;

Digital Machinist (family of sponsors of this forum) had, IIRC, an article on 3D printing on a milling machine about a year ago.

The LinuxCNC group, including Ed Nisley (author of interesting articles on OpenSCAD in said magazine) is one of a bunch trying LinuxCNC to control machines for 3D printing.

I would suggest following up one of the other of the above - Ed can be reached; I have one email for him but am unsure if it is one he wants published - he does have a web site - http://softsolder.com

Might as well see what others have done and decide if that's the direction you want to go?

(me? 3D printing either work or Shapeways - both come out better than I could do at home, but because of building steam locomotives, plastic is not a material of choice!)

Another JohnS

Evan
06-05-2013, 05:19 PM
I have a lot of research to do for sure. I decided to ask here first since there is always somebody on here that knows something about any subject.

ikdor
06-05-2013, 05:20 PM
The feed mechanism is usually just a knurled wheel plus a spring loaded one. You have the advantage you can put it right on top of the extruder as the head doesn't have to fly around. There appears quite some iterations of the extruder. I'd get something like this https://shop.ultimaker.com/en/parts-and-upgrades/v2-hot-end-upgrade.html#
Or find out exactly on the forums why they design the heads like that.

I've always wondered if one could make a crude sintering machine using a tig torch on your CNC with powdered metal......

Igor

Deja Vu
06-05-2013, 06:33 PM
http://cnc2printer3d.wordpress.com/2013/05/31/a-nice-guide-for-cnc-convertion/

Norman Bain
06-05-2013, 07:08 PM
aostling:
Instead of buying one initially I'd like to send a file to a service which could print a part for my inspection. I hope such a service will become available, but I have no knowledge of any.

There are several vendors now offering this service. I expect with the proliferation of home and quasi-commercial units that many will come to offer their expertise and this service generally.

There is a plug-in for most CAD packages designed to send your model to vendors for quoting:
http://www.print3d.com/content/software/print3d-standalone/

Here are a few links I know of:
http://www.shapeways.com/materials/sandstone
http://www.quickparts.com/home.aspx
http://shop.seemecnc.com/

Cheers,
Norman

aostling
06-05-2013, 07:23 PM
There are several vendors now offering this service.

Do you have a link to a service which will print with the Form 1? I need to hold the tolerance which they claim.

Evan
06-05-2013, 07:53 PM
Expect the Form 1 to be expensive. It uses Stereolithography which is the one process I know a little about. It works with a UV scanning head much the same as a laser printer which is something I know. The resins they use are very expensive, in the range of $100 per litre and way up from there. It is one of the most accurate processes because of the ability to focus laser light to a very small spot. It was also the earliest process so has the most development behind it.

kf2qd
06-05-2013, 08:05 PM
Was looking at an article at work today and they had one they mentioned, I looked at the website and the extruder head was $95. Just need a way to control it.

aostling
06-05-2013, 08:34 PM
Expect the Form 1 to be expensive. It uses Stereolithography which is the one process I know a little about.

I think the Form 1 is about $3,200. That's expensive, surely, but cheaper than any machine which can match its capabilities. I might be able to justify the expense for prototyping, but I'm disinclined to take the plunge without seeing some results "in hand." The $100/liter cost of the resin is okay for making a few small parts, but it would be hard to justify that for anything like a production run.

I expect I will wait until a Form 1 printing service is available.

mayfieldtm
06-05-2013, 09:00 PM
I was just looking at the "MetalMagma" Hot End just introduced by TrinityLabs.

I like the idea of being able to print a wide variety of materials up to about 400 degrees C., and the price seems reasonable too.

Here is a quote from their web page...

"This means the MetalMagma hot end can print Nylon, Polycarbonate, PLA, ABS, LayWood, PVA, HDPE and just about any type of filament there is currently available on the market and many more to come in the future".

http://trinitylabs.com/products/trinity-metalmagma-all-metal-high-temp-hotend

Doesn't look all that hard to machine one.

I like your idea of using our CNC Mills with an extruder.
I know that the guys like to print at high speed, I think at around 150-300mm/S.
Might have to adopt some sort of heated bed also.

Tom M.

Evan
06-05-2013, 10:11 PM
Fairly high speed is no problem since there are no cutting loads. Acceleration curves may be an issue when moving around a significant amount of tooling mass.

I am having a good look at that conversion link. Looks very interesting.

George Bulliss
06-06-2013, 07:45 AM
Digital Machinist (family of sponsors of this forum) had, IIRC, an article on 3D printing on a milling machine about a year ago.


The article appeared in the Summer 2012 issue and was written by Dan Mauch. In the article, he explains how he setup his home-built CNC mill to do printing. He also has the Mach3 config file that he uses available for download on our site. DM Downloads (http://www.digitalmachinist.net/downloads)

H8Allegheny
06-06-2013, 08:07 AM
The Form1 does look promising for serious home machinist and hobbyist types and its build envelope is larger than even some ABS/PVC MakerBots. The fly in the ointment with this printer is that there is a potential patent infringement lawsuit pending against it by 3D Systems which could potentially delay or even kill its introduction:

http://www.technologyreview.com/news/515071/next-generation-consumer-3-d-printer-arrives-but-a-lawsuit-looms/?utm_campaign=newsletters&utm_source=newsletter-daily-all&utm_medium=email&utm_content=20130524

I have heard rumors of a high end SLA 3D printer (very fine layer thickness and much larger build volume) that will soon be introduced at a price point of around $8k (vs. the Form1 at $3.2K), but have yet to confirm it.

Rustybolt
06-06-2013, 08:30 AM
I wonder what would happen if you used a MIG welder to build up metal with tiny "spot welds", using CNC timing and position control?

I see that this was mentioned here before but with only one reply it didn't go far. http://bbs.homeshopmachinist.net/threads/55863-Tig-welder-3D-printer

I don't want to sidetrack my own thread though. If anybody has knowledge to share about plastic printing I am all "ears".


Google "Skiaky" They are doing work with metal deposition printing. Apparently there is clean up machining involved, but it looks promising.
I've often wondered if a powerful laser and powdered metal might be the way to go.

MrFluffy
06-06-2013, 08:46 AM
http://cnc2printer3d.wordpress.com/2013/05/31/a-nice-guide-for-cnc-convertion/
That link is great. I already thought about using my up and coming cnc mill project as a cnc plasma with a arm bolted to the table carrying the torch off to one side, but this is a bigger natural extension of its capabilities also.

Dr Stan
06-06-2013, 09:02 AM
Because of its high resolution at (relatively) low cost, I've been waiting for the Form 1 to become widely available http://formlabs.com/products/our-printer. Instead of buying one initially I'd like to send a file to a service which could print a part for my inspection. I hope such a service will become available, but I have no knowledge of any.

Stratasys offers this very service. You can check it out on their web site.

Dr Stan
06-06-2013, 09:06 AM
NASA just spent $125,000 for a study of a food printer. Guess you could crank out a pizza on the mill if you were hungry. :rolleyes:

http://www.scientificamerican.com/article.cfm?id=why-nasa-just-spent-125000-to-fund-2013-06&WT.mc_id=SA_CAT_TECH_20130604

Evan
06-06-2013, 10:23 AM
But can it do "Tea, Earl Grey"?

aostling
06-06-2013, 10:57 AM
Stratasys offers this very service. You can check it out on their web site.

I was unclear. I want a service which will print on the Form 1, to evaluate its suitability before buying one.

Dr Stan
06-06-2013, 11:42 AM
I was unclear. I want a service which will print on the Form 1, to evaluate its suitability before buying one.

Try Emachineshop.

Dr Stan
06-06-2013, 11:43 AM
But can it do "Tea, Earl Grey"?

That will probably be on the phase 2 machine. :)

aostling
06-06-2013, 01:31 PM
I have just started reading this book, which I got from the Phoenix Library: http://www.amazon.com/Fabricated-The-New-World-Printing/dp/1118350634/ref=sr_1_1?ie=UTF8&qid=1370538822&sr=8-1&keywords=fabricated+the+new+world+of+3d+printing. It has much useful information (though not much on stereolithography). It calls for a new file format, stating that STL, which got its start in the 1980s, is now a limitation on the future development of 3D printing.


One possible way to replace and upgrade the STL file is with a new XML-based standard, the Additive Manufacturing Format (AMF). Full disclosure: I co-authored the AMF standard so, of course, I am a fan. I worked on the AMF standard with a group of 3D printer manufacturers, CAD software vendors, and expert users. We teamed up under an international organization that manages the development and implementation of technology standards, the ASTM.

AMF maintains the surface mesh structure of the STL format but has added capabilities to reflect advances in design software and 3D printers. For example, the AMF file format can handle different colors, different types of materials, the creation of lattices, and other detailed internal structures that are one of the huge benefits of additive manufacturing. Curved triangles can be used to describe curved surfaces more accurately and more completely than the planar triangles used in STL.

The AMF standard was officially approved by the standards body in May 2010, but the ultimate test of any standard is its adoption. At the time of writing of this book, it has yet to be adopted by 3D printing vendors. It may take years. We're stuck in a chicken-and-egg paradox: CAD vendors and 3D printing companies are waiting to see whether anyone will gamble on the new format and abandon the old but tired STL.

If your library has the book, I recommend it. You can look inside the book on the Amazon link, and see the contents and some of its pages.

Evan
06-06-2013, 01:57 PM
STL is intended to be a "shrink wrap" file format. Current machines require files to be free of "holes" in the mesh of points that define the surface of the object. This is so the entire outer envelope can then be "sliced" into layers that have entirely closed loops that define all portions, the interior of which must be filled with material. If any loop is open then the program fails.

The requirement to close all loops imposes constraints on what may be printed. As an example it is common to represent a flat surface as a two dimensional plane in CAD programs. If sliced in any direction it either produces a null result or a line with no closed loop. The slicing software takes this into account in the Z axis as each slice will automatically be closed according to slice thickness. It has no way of knowing from the model how thick to make the slice in the X or Y dimension if it isn't defined in the model. The STL format is not designed to carry this information which is the primary limitation. It must be explicitly coded into the model.

It is also necessary to take into account the slice thickness that will be used in the printer. If a flat surface in the model is very close but not quite parallel to the X,Y plane it may result in sequential slices have a very large overlap which at the least will be very time consuming and unnecessary to print.

I am not familiar with the various post processing programs but I am sure they take these factors into account to varying degrees. A new file format would be able to carry this meta information and eliminate much guessing in post processing.

JohnAlex141r
06-06-2013, 02:06 PM
aostling;

(thank you for the link, btw. Had not seen that book)

While I totally agree that the STL format is hideous, surface subdivision is quite a complex issue.

Some format with good support for NURBS (non-uniform rational B-Splines), like the ISO-certified X3D standard (disclosure here - I participated in the standard and conformance testing) is a valid choice. It is used by at least one of the on-line 3D printer companies as the "convert everything to" standard, because it just works for 3D printing. (disclosure here - the technical director of this company - (the 3D goto printer company for the Internet), and I spent a lot of time participating in the same standards and conformance testing work)

Surface subdivision for, say, Hollywood Animation is one thing, but when you are producing very accurate parts, you want to know that what the designer designs and what the machine machines have the same math behind them.


One thing for sure - the nice thing about standards is that there are many to choose from!

Another JohnS.

Evan
06-06-2013, 02:09 PM
We're stuck in a chicken-and-egg paradox: CAD vendors and 3D printing companies are waiting to see whether anyone will gamble on the new format and abandon the old but tired STL.

Silly comment. Nobody will abandon the STL format. They will simply add capability by adding the new format. The main problem with implementing the new format is that not all the possible meta information it may carry is defined yet, such as material types and properties.

JohnAlex141r
06-06-2013, 04:42 PM
STL is intended to be a "shrink wrap" file format. Current machines require files to be free of "holes" in the mesh of points that define the surface of the object. This is so the entire outer envelope can then be "sliced" into layers that have entirely closed loops that define all portions, the interior of which must be filled with material. If any loop is open then the program fails.
...


Hi Evan - just fyi, you'll see these terms used, and my Android App on the play store checks for:

- "watertight" meaning no holes in the shape.
- "2-Manifold" meaning that there are no internal or external "extras".

There should be some pictures of the above on http://freex3d.org in the examples section.

Triangles have zero thickness. The printer programs (or, CAM programs for subtractive machining) "know" how much thickness or whatever, to add.

Since triangles are zero thickness, "normals" (usually in lighting calculations) are used to determine what is the front and back of a triangle. 3D graphics quite often turn off rendering of back-facing triangles for optimization reasons, so sometimes the triangles are only visible from the front.

Hope this is of interest to someone!

Another JohnS.

John Stevenson
06-06-2013, 05:36 PM
But can it do "Tea, Earl Grey"?

You need the "Boston Export" model for that :D

Evan
06-06-2013, 08:41 PM
Hi Evan - just fyi, you'll see these terms used, and my Android App on the play store checks for:

I am very familiar with all the terms as I don't just play with SketchUp. I am fairly familiar with Blender as well as Meshlab, in particular figuring out various formulae to make Meshlab do what I want when filtering meshes. That is essential when reducing polygon counts on 20 gigabyte NASA Digital Terrain Models.

BTW, there is a new plugin available for SketchUp that will shrink wrap models and produce g-code directly within SU. I haven't taken the time to play with it yet to see how well it works.

When it comes to normals that is something many people are not aware of when using SketchUp. SketchUp does care and has facilities to both display and correct inverted faces. SketchUp creates quads instead of triangles but there are several plugins to triangulate all quads. There are also several plugins to export STL files that do it automatically. The newest version of Sketchup that was just released includes STL export in the free version. It is called SketchUp Make.

aostling
06-06-2013, 09:37 PM
BTW, there is a new plugin available for SketchUp that will shrink wrap models and produce g-code directly within SU. I haven't taken the time to play with it yet to see how well it works.


Is this the same task as patching holes in a model, for correcting STL files for 3D printing? I don't want to spent $590 for SketchUp Pro just do to this, so a plugin for regular SketchUp with this capability would be a real saver.

TheAndroid
06-07-2013, 10:31 AM
Evan,
I have a Makerbot and have been working with it to develop prototypes and models. As previously stated, it uses ABS which can warp. However, a bit of hairspray solves that issue.
Undercuts and holes are common in models and the latest skein software (what slices the model into print layers) handle them quite effectively.

aostling
06-07-2013, 12:08 PM
Silly comment. Nobody will abandon the STL format.

The book http://www.amazon.com/Fabricated-The-New-World-Printing/dp/1118350634/ref=sr_1_1?ie=UTF8&qid=1370620831&sr=8-1&keywords=fabricated+the+new+world+of+3d+printing is not technical enough for me to recommend that anybody buy it. If you can get a library copy it is worth reading for its survey of 3D printing history, and discussion of what is likely to come. Your Aunt Fanny could read this book, and not be too confused.

Evan
06-07-2013, 12:22 PM
Is this the same task as patching holes in a model, for correcting STL files for 3D printing?

Short answer=Yes.

The plugin is free and works on Mac. http://www.cadspan.com/tools

Evan
06-07-2013, 12:48 PM
This one also will prepare a model for printing and has no limitations such as the CadSpan plugin. Anything by Tig is very good. He is a very good programmer. You will have to sign up with SketchUcation if you haven't already, which is free.

http://sketchucation.com/forums/viewtopic.php?f=323&t=25466

Another one: http://rhin.crai.archi.fr/rld/plugin_details.php?id=913

Here is another for testing models: http://sketchucation.com/forums/viewtopic.php?t=30504

aostling
06-07-2013, 01:03 PM
The plugin is free and works on Mac. http://www.cadspan.com/tools

Evan,

What a great find. This saves big bucks!

Evan
06-07-2013, 01:28 PM
Make sure you check my following post in case this is on a new page.

aostling
06-07-2013, 02:14 PM
Make sure you check my following post in case this is on a new page.

In fact I did miss your post #40, so thanks again.

aostling
06-07-2013, 07:35 PM
The book mentions this guy, who makes heavy-duty salad bowls by solar 3D printing with a large Fresnel lens. I really did not know it was this simple to melt sand.

http://www.youtube.com/watch?v=Tsk-24UYFs0

The author says this could be a solution to making roads in sandy areas, using a traveling solar lens to print the road by melting sand. Loose sand blown onto the road would be sintered in subsequent passes, until a sufficient thickness was built up. The procedure would be entirely solar powered.

lazlo
06-07-2013, 10:19 PM
http://formlabs.com/products/our-printer.

The Form 1 is wildly behind schedule. They pissed off a bunch of Kickstarter funders who were promised machines October 2012. Ever since, their status has been shipping "by the end of next week."

I think they're the subject of promising too much for too little: they collected $2 million dollars, 20 times their goal, in the first month they were on Kickstarter. To put that in perspective, that's 1,000 Form 1's they're supposed to deliver to early funders, in the first month they were on Kickstarter :rolleyes:

Evan: someone posted a video of a commercial MIG-based 3D printer here, but I can't find it. It was making a truss structure. Pretty primitive/low resolution, even compared to the ABS plastic home machines.

lazlo
06-07-2013, 10:30 PM
The book mentions this guy, who makes heavy-duty salad bowls by solar 3D printing with a large Fresnel lens. I really did not know it was this simple to melt sand.

http://www.youtube.com/watch?v=Tsk-24UYFs0.

That's pretty slick, but that machine suffers the bane of all 3D printing: each layer you sinter you have to pull a layer of material over, start the sinter, level it off, repeat many, many times. I bet that bowl took all day.

aostling
06-08-2013, 12:10 AM
That's pretty slick, but that machine suffers the bane of all 3D printing: each layer you sinter you have to pull a layer of material over, start the sinter, level it off, repeat many, many times.

The inventor, dragging his machine into the first frame of the video, is channeling Sisyphus, hamming it up. But the concept does give rise to some blue-sky thinking.

The notion of a solar-sintering road builder, for example, suggests ways to do some planetary-engineering on the planet Mercury. You could have parabolic cylindrical Fresnel lens focusing the intense sun's rays to a line focus instead of a point, a line as long as a road is wide. There could be an array of these melt zones, one after the other, with a device like a flour sifter laying down sand automatically for multiple layers.

Well, would could you do with a road on Mercury? Better to cast some above-ground tunnels.

Evan
06-08-2013, 01:18 AM
I have been doing a lot of looking at various models and designs. I am underwhelmed, to say the least. The plastic deposition printers in general turn out pretty useless parts that do not look attractive or at all precise. The printers themselves are mostly quite crude as are the various parts from which they are made. Even the so called "high quality" parts don't impress me much. I have seen almost nothing that I would consider high quality.

There is certainly room for improvement, a great deal of improvement. I can think of a number of ways to improve the extruder and heater components alone. Resolution is terrible with effective resolution on most plastic machines being maybe around 20 to 30 thou regardless of what is specified. I also see people making many items that would be far better made for much less time and money on a really cheap milling machine. Since they are making things from plastic it isn't much of a challenge to make on just about any mill on the market. Many of these items are being made simply because then they can say it was "3D printed". At this point it strikes me as a dancing bear; It isn't how well it dances, it's the fact that it dances at all.

small.planes
06-08-2013, 04:02 AM
And yet in my garage is a 3d printer that its built from precision shafting and parts printed on another 3d printer which assembled with NO fettling of the parts at all.

I have used various rapid prototyping technologies whilst at Phase Vision, and the right ones are capable of turning out commercial quality parts in very short lead times with very few restrictions on shape or complexity. You have to think differently, and certainly the commercial machines are more expensive than home built, but the other night I printed a Lego brick for kicks.it clips into regular Lego just fine.
Seems in this area you are not yet a world expert...

Dave

Evan
06-08-2013, 10:43 AM
I most certainly am not expert. What I am commenting on is the plastic deposition models using plastic filament with usually .3 mm extruders and at best .1 mm layers. While .1mm layers (~.004") seems fairly good the results belie that impression. At this point in my investigation of this process I fail to see the utility of this type of additive machining. It appears to be mostly a toy.

JohnAlex141r
06-08-2013, 12:26 PM
...At this point in my investigation of this process I fail to see the utility of this type of additive machining. It appears to be mostly a toy.

Whilst I personally agree, I have been following (commercial interests not for discussion here) and what I have seen since my first introduction to the original Makerbot "cupcake" printer (SIGGRAPH LA - 2008 or 2010), to what is available now is interesting.

There's no comparison. Give it another 5 years, and we'll see.

Are the current crop of printers "toys"? Is a Unimat SL lathe a "toy"? Not if it does what you want, then it's a tool.

Actually, for the first time machining, I made a bracket out of aluminium and wished I had a 3D printer to print it out. Whilst I do have limited access to a commercial 3D printer, the aluminium bit will do, of course.

Another JohnS.

John Stevenson
06-08-2013, 12:59 PM
At this point in my investigation of this process I fail to see the utility of this type of additive machining. It appears to be mostly a toy.


Sorry Evan but I strongly disagree with you.

Take this motor.

http://www.jprelec.co.uk/image/450-005_big.jpg


I have just grabbed this image off the net and it's not one of my motors but it's very generic to ones I come across but larger.

Often they have aluminium ends and phenol brush rings and when they go pear shaped they set fire to the rings.
simple job with a CNC router to make new rings.

However when these motors of a type in the picture catch fire they are toast. Now if it's simple mass produced motor, no sweat bin it and get a new one.
However if it's a special they can get quite spendy. I recently had a fan motor rewound out of a control cabinet off a big 7 Kw laser cutter, it cost 250 for the rewind for a motor what would fit into the palm of your hand.
Siemens wanted 900 for a new one.

Just a couple of motor ends would pay for a Rep-rap machine.

When I first bought my laser cutter I worked out it that it was 90% toy for John and 10% work horse, that was enough to justify the purchase ;)
Nowadays it's 70% work horse and 30% toy and it's getting better as the word gets round for gaskets and bits.
Got an order last week for 100 cases similar to a raspberry pi case for some special circuit boards.

aostling
06-08-2013, 01:09 PM
.. the other night I printed a Lego brick for kicks.it clips into regular Lego just fine.


Each of the planar and cylindrical surfaces of a Lego block are parallel to one of the X,Y, or Z axes of the printer. There are no surfaces on which a "jaggy" could occur, so the Lego is not a very challenging test piece. How smoothly does your printer produce, say, a hemispherical cap?

small.planes
06-08-2013, 01:29 PM
Yep you are right, but its only been running for a couple of days. I'm still getting the CA.D side of it sorted as the garage PC its ancient, and the inside one is running a beta of 123d with some scaling issues. I was just happy it c came out the size that Iexpected...
It does triangle circle square test very well, triangle pyramid is to come
.Of course what people have also missed is that printed part doesn't have to be used directly. Its perfectly possible to print +1mm and then machine to final size. Called near net production or something like that.

Dave

John Stevenson
06-08-2013, 01:35 PM
Get kettle on when i get back, I'll bring some STL files and some decent coffee and you can impress me. ;)

Evan
06-08-2013, 01:38 PM
Sorry John. There is no comparison between a laser cutter and a RepRap machine. Please pay attention to what I am talking about. It is not about 3D printing in general. It's about the specific types of machines I mentioned above. As soon as you move away from those machines the technology and especially the costs go sky high. Naturally, the quality goes up considerably.

The problem is that the hobbyist machines are more toys than usable additions to the shop. I am interested in the method in general but it needs to be something I can actually use to supplement my shop capability. I can already make things from plastic of high complexity so the only thing 3D printing has to offer is the type of geometry not attainable via subtractive machining.

What I am now investigating is if the plastic process can be improved to the point that it can produce really useable parts to a reasonable standard of accuracy. What I find so far is "gears" that I could make better using a hack saw and file. Part of the problem is that it seems the majority of people using these low end machines have no acquaintance with actual machining.

small.planes
06-08-2013, 01:47 PM
Part of the problem is that it seems the majority of people using these low end machines have no acquaintance with actual machining.
I think you are probably correct. This is more of a 'technogeek' hobby, where mostly it is software people. In the quite recent past you needed the software background just to get the things running...


What I find so far is "gears" that I could make better using a hack saw and file.

You are better with a file than I am (or have more patience) :http://thingiverse-production.s3.amazonaws.com/renders/fe/5c/2d/7d/d2/003_display_large.jpg

Dave

lazlo
06-08-2013, 01:50 PM
I have been doing a lot of looking at various models and designs. I am underwhelmed, to say the least.

Evan, I haven't been following the various OpenSource projects (they've got a long way to go before they're more than toys, as you say), but I did see a blog article from the Mendel group where one of the team members built a 3rd gen extruder head that looked a lot better designed than the usual RepRap/MakerBot fare.

I'll see if I can find it.

lazlo
06-08-2013, 01:54 PM
I think you are probably correct. This is more of a 'technogeek' hobby, where mostly it is software people.

You are better with a file than I am (or have more patience) :

I admit -- I'm impressed! :) I've seen lots of lousy prints of that ThingVerse "Screwless Heart Gear", but that looks pretty nice. What extruder head is that?


And yet in my garage is a 3d printer that its built from precision shafting and parts printed on another 3d printer which assembled with NO fettling of the parts at all.

Can you post some pictures of the machine you've built??

John Stevenson
06-08-2013, 01:58 PM
I wasn't comparing a laser cutter to 3D printer, I was paying attention, when you have been married for as long as I have you can't get away with NOT paying attention.

The point I was making is that the chargeable time on the laser now is 60% up from what I envisaged and I can see a 3D printer doing the same IF you can spot niche markets.

I can also make plastic parts of high complexability as well because we both have similar equipment. However the 3D printer has two advantages, one is that I don't have to turn handles as much, if at all and secondly you can insert metal objects into the part as it's being made.

Imagine things like rectangular brass brush boxes being embedded into the extrusion as it's growing, or a dumbbell shaped terminal block so it can't pull thru.

John Stevenson
06-08-2013, 02:07 PM
Evan, I haven't been following the various OpenSource projects (they've got a long way to go before they're more than toys, as you say), but I did see a blog article from the Mendel group where one of the team members built a 3rd gen extruder head that looked a lot better designed than the usual RepRap/MakerBot fare.

I'll see if I can find it.

Robert,
It's always easier to improve on a design than to come up with a complete new design.
As Dave has said it very geek driven and simple enough that it can be built on the kitchen table with a screw driver and can opener.

The next step is for someone with a flair for design to take say the extruder head and think If we make this part out of this, change this to this and before long you have an industrial pattern head that's repeatable in a workshop but not the kitchen table.

We need innovation and later on perfection. I don't know whether to call it innofection or pervation ?
Think I'll stick to bodging ;)

small.planes
06-08-2013, 02:30 PM
Ive not got many pics at the mo, but its a reprappro Huxley.
http://i306.photobucket.com/albums/nn274/small_planes/Huxley/DSC_1162_zps49a3fc99.jpg

Ive not tweaked any of the out of the box settings yet, but this is 0.5mm wall thickness, 10mm high:
http://i306.photobucket.com/albums/nn274/small_planes/Huxley/DSC_1167_zps70632b0f.jpg

Dave

Jpfalt
06-08-2013, 06:54 PM
I own two of the dimension 3D printers. One is an SST768 and the other is an SST1200es. All the pieces are out there for totally useable parts. I have been printing functioning prototypes for about 8 years now. These included functioning string trimmer heads that ran at 10,000 rpm and an 8" blower impeller for a backpack blower that ran at about 7500 rpm. I added a single lockwire loop around the blower impeller to prevent it from stretching in use. I have also done a lot of master patterns for making castings and have even used parts with a sparse interior for investment casting. They burn out of the mold pretty well.

The main issues I see with the open source machines is the working envelope. Several layers of deposit do the same thing that weld beads do. They progressively pull the layers underneath them into a reverse arch. Dimension deals with this by having the working envelope at elevated temperature to keep the layers underneath warm enough that the warp is minimized, but cool enough that the plastic will harden enough quickly enough to hold shape.

The extruder heads are fairly simple, just a temperature controlled heater, a tube and a stepper motor with capstan to feed the plastic filament.

I think the second biggest issue that needs to be resolved is software for handling PLA soluble support under either ABS or styrene. To me, that is the most difficult issue if you are going with fusion deposition method.

Evan
06-08-2013, 06:55 PM
What I am seeing here looks reasonable but it is probably near the top of the heap too. I have been considering some of my own requirements and one will be at least three extruder heads and preferably five. That will allow for the three primary colours plus white and black to be selected at will.

I also would like to use acetal or maybe nylon rather than ABS or PLA. Acetal and Nylon are much higher on the structural plastic scale. Also is PETG which is in the same league as polycarbonate without the bubbles. It also has a very wide liquid temp range with viscosity directly proportional over that range. It should be extremely controllable for extruding as well as being crystal clear which allows for nice looking artwork. Art is one area where the precision is not nearly as important and it is also an interest of mine. I have messed about with plastics at melting temperatures including thermoforming and vacuforming so I have a good idea of what works and how.

Robert, anything you can find that is better than what I am mostly seeing if of interest to me. What I have not been able to find is anyone extruding at .1 mm nozzle size.

Jpfalt
06-08-2013, 07:05 PM
Nylon will be a lot more difficult to use for FDM. The melting point is enough higher than ABS, PLA or styrene to make the envelope temperature a greater dilemma. If the temperature is kept down, the parts will warp badly. If the temperature is high enough to control the warpage, then the components in the head have to be specifically made for the high temperatures. The other disadvantage to using nylon is reduced strength in the vertical direction. ABS and styrene can be fused with acetone to make them equally strong in all three directions whild nylon cannot be conveniently fused in the vertical direction.

Evan
06-08-2013, 07:18 PM
What I have been reading about nylon is that it fuses extremely well. As for the temperature, that just means not using plastics in the extruder. I wouldn't anyway as there are plenty of suitable metal alternatives if you have a machine shop. One guy has some videos on Utube building using string trimmer nylon line. Seems to work at least as well as the other plastics.

small.planes
06-09-2013, 04:37 AM
People are using nylon. Be aware that it needs baking before use to remove that water so it extrudes nicelu.

An all metal head and carriage is quite simple for a hsm type.

Dave

camdigger
06-09-2013, 04:59 AM
http://blog.makezine.com/volume/make-ultimate-guide-to-3d-printing/?utm_source=Contextly&utm_medium=RelatedLinks&utm_campaign=Previous

A small investment in some reading material from MAKE...

camdigger
06-09-2013, 05:01 AM
http://blog.makezine.com/volume/make-ultimate-guide-to-3d-printing/?utm_source=Contextly&utm_medium=RelatedLinks&utm_campaign=Previous

A comparison of available printers from MAKE. Seems a paltry investment to get some info from people who've been there, done it, and are willing to share what they learned.

vincemulhollon
06-09-2013, 07:52 AM
....but because of building steam locomotives, plastic is not a material of choice...

Ah don't write it off so quick. On my infinite list of things to do is get a large enough 3-d printer to print out usable foundry patterns. I really wanna try this and I can't find any examples of folks with "home grade" 3-d printers already doing this, although with all the complaining about ABS and PLA it seems obvious, to me anyway, to 3-d print foundry stuff.

There's a couple centuries experience with wood butchers making relatively simple patterns, because, after all, time is money. But what if you could cast pretty much anything of arbitrary complexity with near perfect repeatability? I can't immediately think of anything, but someone will eventually.

Of course if you thought shrinkage and warpage was a problem with just 3-d printing or just foundry work, imagine twice as much back to back for extra fun.

Combined with some kind of open source-ish workflow, it would be interesting to share perfect foundry patterns. Here is revision 2352 of something like a Gingery lathe bed with absolutely perfect shrinkage allowances and absolutely perfect draft etc. Given a thermocouple to verify your pour temperature is perfect, reasonably standard alloy (aluminum piston-ium?) and a relatively standardized sand (bog standard otta the bag premix petrobond, perhaps?), and some standard for prep work (maybe a video showing exactly how the designer pounds the sand into place?) I think near perfect castings would almost inevitably result, every time.

Like a factory on its 2000th production pour, except distributed all around the planet instead of just one factory.

JohnAlex141r
06-09-2013, 11:20 AM
...Ah don't write it off so quick. On my infinite list of things to do is get a large enough 3-d printer to print out usable foundry patterns....


VinceMuhollon;

Very true about the patterns - somewhere I have seen home-3d printed patterns used for Aluminium pours, so it has been proven true.

I guess, in *my* workshop, I'm trying to fabricate everything, including wheel "castings" for live steam locomotives.

Pattern making with 3D is a very good use, as you say. I should have been more specific - my apologies.

Also, some club members are getting things done in 3D printed "bronze", which seems to be turning out (no pun intended) well.

This is a very interesting thread - thank you all for your thoughts and ideas.

Another JohnS.

lazlo
06-09-2013, 12:05 PM
http://blog.makezine.com/volume/make-ultimate-guide-to-3d-printing/?utm_source=Contextly&utm_medium=RelatedLinks&utm_campaign=Previous

A comparison of available printers from MAKE. Seems a paltry investment to get some info from people who've been there, done it, and are willing to share what they learned.

I was about to send Evan a link to the PDF of that article :)

It's not very good. For one thing, Make has their own machines in the lineup, so it's hardly an unbiased review. But it's a good overview of the current state of the art for purchased (not Open Source) amateur machines.

lazlo
06-09-2013, 12:36 PM
Evan: just sent you a PM.

Based on a quick scan last night, the "3D Touch" seems like the most mechanically robust design in this price point, and would be a straightforward build for most here.

Another robust design is the $4,000 Aluminatus printer. It's made from off-the-shelf Bosch extrusions and linear guides:

http://trinitylabs.com/pages/aluminatus-overview

Neither is Open Source, so you'd have to reverse engineer it.

I can't find the worm-gear extruder I had seen last year from one of the Mendel guys, but the current state of the art seems to be a planetary gear-head stepper driving a toothed drive wheel like the Hyena 2.0. People are mixing and matching the "hot end" (the nozzle):

http://blog.arcol.hu/
http://farm9.static.flickr.com/8150/7183878128_ee1ff21b05.jpg

Evan
06-09-2013, 01:24 PM
Thanks Robert. I just downloaded it and will read it later.

Everyone please note again I have NOT said that 3D printing is useless or just a toy. Quite the contrary. But, I do not see much use for the current crop of plastic deposition printers and I suspect they are close to the limits of resolution that can be reasonably achieved by that process as it is currently implemented. There are many other ways to do 3D printing and all of them are expensive at this time. One reason for that is the pricing is currently at the "early adopter" phase where profits are very high and production runs are small. This is even true of the plastic printers where ABS 1 kilo spools are selling with around 100% markup over advertised prices for 100 kilo lots.

I just checked bulk ABS resin pricing as of last Friday and it is running from 115 to 118 cents per pound. edit: A 2.2 lb spool of ABS is currently selling from about $30 to $50.